From: Jeff King <peff@peff.net>
Date: Tue, 13 Jan 2015 19:58:06 GMT
Subject: Re: [PATCH] http-push: trim trailing newline from remote symref
Message-ID: <20150113195806.GA13282@peff.net>
In-Reply-To: <4E1001F7-2D26-483A-9339-E2570559ECF0@gmail.com>

On Tue, Jan 13, 2015 at 08:26:31AM -0800, Kyle J. McKay wrote:

> I have this line in my 2.1.4 test output log:
> 
> t5540-http-push-webdav.sh .......................... ok
> [...]
> I do not build with NO_EXPAT.  This is running the tests on OS X without
> this patch applied.  Is something else required to get a failure?

Hmm. I think it is probably this:

  http://curl.haxx.se/docs/adv_20150108B.html

where curl has started complaining about URLs with newlines in them. So
ae021d8 did introduce a bug, but older versions of curl did not really
care. The combination of ae021d8 with a new version of curl triggers the
problem.

And that also explains why it worked prior to eecc8367f4; curl was more
forgiving. Also, interestingly, if you "git log -S'- 6' http-push.c",
you can see the exact same bug reappear and go away in 2006/2007. The
implicit "chop one character" behavior is there in the original
3dfaf7bc, adding http-push support. Then it disappears as a side effect
of bfbd0bb6, and then comes back again in eecc8367.

Anyway. I think my patch is still the right thing. But that does explain
why we didn't notice the test failure.

-Peff